Probably a bug in enable pin management

Home Forums gShield grblShield Support Probably a bug in enable pin management

Viewing 2 posts - 1 through 2 (of 2 total)
  • Author
    Posts
  • #1631
    ewidance
    Member

    Hi,

    I’ve used your grbl patched version with a TB66560 driver (not grblshield… Sorry). Since my tb6560 fully works with positive logic, I’ve done proper signal inversion on enable in stepper.c (st_wake_up(), st_disable_steppers()) and reset DEFAULT_STEPPING_INVERT_MASK.

    Works but… It seems when using the enable pin, motors loose steps. When running many times G0x100 G0X00, motors DOES NOT returns to initial position. Motors are not driven too fast, and acceleration is properly setup.

    It does not seems to be proportional to move menght (ex : G0x1000 G0X00), but more on number of moves.

    I’ve investigated a bit and found if i’m hardware wiring my TB6560 enable signal to 5v instead of the pin8 or arduino, there’s no more step loose.

    I suspect a timing problem with enable signal. I don’t know for the moment if the signal is asserted too late or unasserted to quickly.

    Any idea?

    #1632
    alden
    Member

    Thanks for he heads up. Sounds like you have a solution to your problem, however. Since you are in there at the grbl source code level I’d just enable the enable pin all the time. I can’t really answer the timing question you ask about the innards of the grbl code, but you might post this as an issue to the grbl github.

    Also, since you are not using grblshield you may just want to get the “official” grbl release from the grbl github (simen/grbl). The changes in the grblshield patched version are there to support the pin signal polarities in grblshield and since you don’t need that I don’t think there’s any advantage to using the patched version.

    By the way, the grbl edge branch (grbl version 0.7) is now compatible with grblshield, so we won’t be needing the patched version once the grbl project moves version 0.7 to the master branch (and publishes the master .hex file).

    Thanks for letting us know about this anomaly.

Viewing 2 posts - 1 through 2 (of 2 total)
  • You must be logged in to reply to this topic.